diff --git a/src/cpp.c b/src/cpp.c
index 2317480cd55c862ac48f0342a3b197254e5cab52..9e24c19d28fc337791ea0cc5a1de72e2579bee7d 100644
--- a/src/cpp.c
+++ b/src/cpp.c
@@ -5,7 +5,7 @@
\*/
/*
- * $Id: cpp.c,v 1.40 1999/02/23 02:21:15 grubba Exp $
+ * $Id: cpp.c,v 1.41 1999/02/24 01:50:20 grubba Exp $
*/
#include "global.h"
#include "language.h"
@@ -2080,6 +2080,145 @@ static int do_safe_index_call(struct pike_string *s)
return res;
}
+static struct pike_string *recode_string(struct pike_string *data)
+{
+ fprintf(stderr, "recode_string()\n");
+
+ /* Observations:
+ *
+ * * At least a prefix of two bytes need to be 7bit in a valid
+ * Pike program.
+ *
+ * * NUL isn't valid in a Pike program.
+ */
+ /* Heuristic:
+ *
+ * Index 0 | Index 1 | Interpretation
+ * --------+---------+------------------------------------------
+ * 0 | 0 | 32bit wide string.
+ * 0 | >0 | 16bit Unicode string.
+ * >0 | 0 | 16bit Unicode string reverse byte order.
+ * 0xfe | 0xff | 16bit Unicode string.
+ * 0xff | 0xfe | 16bit Unicode string reverse byte order.
+ * 0x7b | 0x83 | EBCDIC-US ("#c").
+ * 0x7b | 0x40 | EBCDIC-US ("# ").
+ * 0x7b | 0x09 | EBCDIC-US ("#\t").
+ * --------+---------+------------------------------------------
+ * Other | Other | 8bit standard string.
+ *
+ * Note that the tests below are more lenient than the table above.
+ * This shouldn't matter, since the other cases would be erroneus
+ * anyway.
+ */
+
+ /* Add an extra reference to data, since we may return it as is. */
+ add_ref(data);
+
+ if ((!((unsigned char *)data->str)[0]) ||
+ (((unsigned char *)data->str)[0] == 0xfe) ||
+ (((unsigned char *)data->str)[0] == 0xff) ||
+ (!((unsigned char *)data->str)[1])) {
+ /* Unicode */
+ if ((!((unsigned char *)data->str)[0]) &&
+ (!((unsigned char *)data->str)[1])) {
+ /* 32bit Unicode (UCS4) */
+ } else {
+ /* 16bit Unicode */
+ if ((!((unsigned char *)data->str)[1]) ||
+ (((unsigned char *)data->str)[1] == 0xfe)) {
+ /* Reverse Byte-order */
+ }
+ // Note: We lose the extra reference to data here.
+ push_string(data);
+ f_unicode_to_string(1);
+ add_ref(data = sp[-1].u.string);
+ pop_stack();
+ }
+ } else if (((unsigned char *)data->str)[0] == 0x7b) {
+ /* EBCDIC */
+ /* Notes on EBCDIC:
+ *
+ * * EBCDIC conversion needs to first convert the first line
+ * according to EBCDIC-US, and then the rest of the string
+ * according to the encoding specified by the first line.
+ *
+ * * It's an error for a program written in EBCDIC not to
+ * start with a #charset directive.
+ *
+ * Obfuscation note:
+ *
+ * * This still allows the rest of the file to be written in
+ * another encoding than EBCDIC.
+ */
+ }
+ return data;
+}
+
+static struct pike_string *filter_bom(struct pike_string *data)
+{
+ /* More notes:
+ *
+ * * Character 0xfeff (ZERO WIDTH NO-BREAK SPACE = BYTE ORDER MARK = BOM)
+ * needs to be filtered away before processing continues.
+ */
+ int i;
+ int j = 0;
+ int len = data->len;
+ struct string_builder buf;
+
+ fprintf(stderr, "filter_bom()\n");
+
+ /* Add an extra reference to data here, since we may return it as is. */
+ add_ref(data);
+
+ if (!data->size_shift) {
+ return(data);
+ }
+
+ init_string_builder(&buf, data->size_shift);
+ if (data->size_shift == 1) {
+ /* 16 bit string */
+ p_wchar1 *ptr = STR1(data);
+ for(i = 0; i<len; i++) {
+ if (ptr[i] == 0xfeff) {
+ if (i != j) {
+ string_builder_append(&buf, MKPCHARP(ptr + j, 1), i - j);
+ j = i+1;
+ }
+ }
+ }
+ if ((j) && (i != j)) {
+ /* Add the trailing string */
+ string_builder_append(&buf, MKPCHARP(ptr + j, 1), i - j);
+ free_string(data);
+ data = finish_string_builder(&buf);
+ } else {
+ /* String didn't contain 0xfeff */
+ free_string_builder(&buf);
+ }
+ } else {
+ /* 32 bit string */
+ p_wchar2 *ptr = STR2(data);
+ for(i = 0; i<len; i++) {
+ if (ptr[i] == 0xfeff) {
+ if (i != j) {
+ string_builder_append(&buf, MKPCHARP(ptr + j, 2), i - j);
+ j = i+1;
+ }
+ }
+ }
+ if ((j) && (i != j)) {
+ /* Add the trailing string */
+ string_builder_append(&buf, MKPCHARP(ptr + j, 2), i - j);
+ free_string(data);
+ data = finish_string_builder(&buf);
+ } else {
+ /* String didn't contain 0xfeff */
+ free_string_builder(&buf);
+ }
+ }
+ return(data);
+}
void f_cpp(INT32 args)
{
@@ -2111,131 +2250,19 @@ void f_cpp(INT32 args)
this.current_file=make_shared_string("-");
}
- data = sp[-args].u.string;
+ add_ref(data = sp[-args].u.string);
if (auto_convert && (!data->size_shift) && (data->len > 1)) {
/* Try to determine if we need to recode the string */
-
- /* Observations:
- *
- * * At least a prefix of two bytes need to be 7bit in a valid
- * Pike program.
- *
- * * NUL isn't valid in a Pike program.
- */
- /* Heuristic:
- *
- * Index 0 | Index 1 | Interpretation
- * --------+---------+------------------------------------------
- * 0 | 0 | 32bit wide string.
- * 0 | >0 | 16bit Unicode string.
- * >0 | 0 | 16bit Unicode string reverse byte order.
- * 0xfe | 0xff | 16bit Unicode string.
- * 0xff | 0xfe | 16bit Unicode string reverse byte order.
- * 0x7b | 0x83 | EBCDIC-US ("#c").
- * 0x7b | 0x40 | EBCDIC-US ("# ").
- * 0x7b | 0x09 | EBCDIC-US ("#\t").
- * --------+---------+------------------------------------------
- * Other | Other | 8bit standard string.
- *
- * Note that the tests below are more lenient than the table above.
- * This shouldn't matter, since the other cases would be erroneus
- * anyway.
- *
- * Note:
- *
- * * The code below may leave some extra strings on the stack.
- */
- if ((!((unsigned char *)data->str)[0]) ||
- (((unsigned char *)data->str)[0] == 0xfe) ||
- (((unsigned char *)data->str)[0] == 0xff) ||
- (!((unsigned char *)data->str)[1])) {
- /* Unicode */
- if ((!((unsigned char *)data->str)[0]) &&
- (!((unsigned char *)data->str)[1])) {
- /* 32bit Unicode (UCS4) */
- } else {
- /* 16bit Unicode */
- if ((!((unsigned char *)data->str)[1]) ||
- (((unsigned char *)data->str)[1] == 0xfe)) {
- /* Reverse Byte-order */
- }
- push_string(data);
- f_unicode_to_string(1);
- data = sp[-1].u.string;
- }
- } else if (((unsigned char *)data->str)[0] == 0x7b) {
- /* EBCDIC */
- /* Notes on EBCDIC:
- *
- * * EBCDIC conversion needs to first convert the first line
- * according to EBCDIC-US, and then the rest of the string
- * according to the encoding specified by the first line.
- *
- * * It's an error for a program written in EBCDIC not to
- * start with a #charset directive.
- *
- * Obfuscation note:
- *
- * * This still allows the rest of the file to be written in
- * another encoding than EBCDIC.
- */
- }
+ struct pike_string *new_data = recode_string(data);
+ free_string(data);
+ data = new_data;
}
if (data->size_shift) {
- /* More notes:
- *
- * * Character 0xfeff (ZERO WIDTH NO-BREAK SPACE = BYTE ORDER MARK = BOM)
- * needs to be filtered away before processing continues.
- *
- * * The code below may leave some extra strings on the stack.
- */
- int i;
- int j = 0;
- int len = data->len;
-
- init_string_builder(&this.buf, data->size_shift);
- if (data->size_shift == 1) {
- /* 16 bit string */
- p_wchar1 *ptr = STR1(data);
- for(i = 0; i<len; i++) {
- if (ptr[i] == 0xfeff) {
- if (i != j) {
- string_builder_append(&this.buf, MKPCHARP(ptr + j, 1), i - j);
- j = i+1;
- }
- }
- }
- if ((j) && (i != j)) {
- /* Add the trailing string */
- string_builder_append(&this.buf, MKPCHARP(ptr + j, 1), i - j);
- push_string(finish_string_builder(&this.buf));
- data = sp[-1].u.string;
- } else {
- /* String didn't contain 0xfeff */
- free_string_builder(&this.buf);
- }
- } else {
- /* 32 bit string */
- p_wchar2 *ptr = STR2(data);
- for(i = 0; i<len; i++) {
- if (ptr[i] == 0xfeff) {
- if (i != j) {
- string_builder_append(&this.buf, MKPCHARP(ptr + j, 2), i - j);
- j = i+1;
- }
- }
- }
- if ((j) && (i != j)) {
- /* Add the trailing string */
- string_builder_append(&this.buf, MKPCHARP(ptr + j, 2), i - j);
- push_string(finish_string_builder(&this.buf));
- data = sp[-1].u.string;
- } else {
- /* String didn't contain 0xfeff */
- free_string_builder(&this.buf);
- }
- }
+ /* Get rid of any byte order marks (0xfeff) */
+ struct pike_string *new_data = filter_bom(data);
+ free_string(data);
+ data = new_data;
}
init_string_builder(&this.buf, 0);
@@ -2276,6 +2303,8 @@ void f_cpp(INT32 args)
free_string(this.current_file);
+ free_string(data);
+
if(this.compile_errors)
{
free_string_builder(&this.buf);